"If you haven't figured out how to use [AI] by now, it's a signal [to hiring managers] about how you learn" This is quite a statement ... and yes, this is a sub-skeet. Being an AI advocate should be a red flag...

Annoyed that I had to open Chrome yesterday because a plumber's website's basic functionality required a popup that for some reason didn't work except in Chrome.

“My thoughts are precious and being able to form them through writing and socializing with others currently has no equal practice. Prompting appears to give the benefits of both the writing and socializing at once, but it actually steals the benefits of doing both of these acts from us.”

it’s hard not to see the “agentic” future as an illustration of vc’s stunted vision of the future. not more creative freedom, not new access to voices and ideas, rather the opportunity to simulate the true dream of capital: managing an army of makers whose work becomes yours.

This week a 7.1 million-square-foot hyperscale data center was approved for construction about 20 miles from my rural hometown. It will be the 3rd largest in the world and sit adjacent to the Big Oaks National Wildlife Refuge. None of this makes me happy.

This is such a clever idea! Haven't gotten it yet, but will after work today!

Totally agree. They’re going to sell making your website ready for agents as the new SEO. But then no one will have to visit your site. I used to love the web. It was useful, exciting, creative, a community, and I could get paid for joining in. That’s feeling impossible now. Fuck Google. Fuck AI.

I wanted to see if anything had obviously changed in search since I/O announcements (which I don't want to watch), so I searched "test" and this is the worst type of thing to spend billions on... This industry is devoid of meaning at this point. No one should support this technology

Imagine if instead of spending billions on LLMs that no one wants, Google had spent it on creating an education ecosystem where developer educators could safely spend their lives teaching the next generation of developers everything from the basics to niche skills

There's a difference between making a good product and making a product that generates shareholder value. Sometimes they cross over, but if you focus on the latter, you won't make the former

Sir, another graduation ceremony in which the students booed AI has dropped. "College graduates were pissed after their school used AI to announce graduates’ names and missed hundreds of names" (via @/FearedBuck on Twitter)

Very frustrating that Google "summarizes" my opinions when asked, and then mis-titles my YouTube series and doesn't provide a link to it even in the small "linked" area ... which, let's be honest, isn't incentivizing clicks

Great write-up on the less-talked-about human cost of LLMs in software development. There's a dehumanization at play with all of these tools. It's enough to not feel accidental. It's enough to make me question and wonder if breaking communal bonds wasn't the intention all along. Community = power

Managing the rage that AI workflow pushers give me is becoming much harder to deal with on a daily basis

In what may be my most ambitious plan yet for the @whenweplay.org after-school program, we'll be playing 2 tables of King of Tokyo this Friday. I'm relatively confident these kids can handle it, but... not 100%. Upside, I'll bring short games that we've already learned for if kids get knocked out

I see people who are pro-LLMs that are also like “I wouldn’t trust a college degree anymore because of LLMs” and they just keep boosting. It’s a systemic problem. If you see the problem (well, one problem), why boost it; why give these companies money and cover?!

So much terrible in this, but let me pull this quote: "Agents will plan, code, review, deploy, and repair." For everyone that says "there will be a human in the loop," I wonder... where's the human in that loop?

As always, "AI" is a technology about the breaking down of consent by the powerful. In this one move alone we have lack of consent for browser users, lack of consent from the developer ecosystem, and lack of consent from the web standards groups. Just push ahead anyway

At this point, I’m fed up with telling people why I think using GAI to read write or think is bad. Fine, then. Just tell me when you use it. Put a label on gAI writing. So I know to avoid it. But you won’t do that either, will you? Because you KNOW what you’re doing is sketchy.

Dang. www.stvn.sh/writing/prog...

New blog post for @whenweplay.org: Set is a near-perfect game for building brains, learning resiliency, and becoming patient learners. whenweplay.org/blog/set-per...

This week's after-school program for @whenweplay.org will involve new school and old school. Word games and abstract games. Word Games: - Boggle - Bananagrams Abstract Games: - Pentago (imagine if tic-tac-toe or Connect4 was good) - boop! A fantastic, cat-themed abstract game

sorry to christopher nolan but this is the best adaptation of the odyssey

Running events at "purpose-built" places is interesting. Get folks going to those places to relax and chill isn't easy. At our gym/rec center last night, watching the bustle. Everyone is highly focused on the task they came for and almost have blinders on.

🔗 Another reason to disallow LLM contributions in open source projects would be that it gets in the way of the project nurturing “new, confident, trustworthy contributors“, says (even) Simon Willison simonwillison.net/2026/Apr/30/... (via Cassandra on the Fediverse)

Actual “AI” literacy would involve explaining why it's not and never will be a reliable source of information nor will it ever be a replacement for research or creativity.

God they’re doing this shit to kids too.

It gives me no joy to be a broken record and a Cassandra about how much this shit is a cudgel against not just the working class, but anyone who they feel as though they can replace with synthetic text, in order to straight up rob them of the wages they are due.

Last week in our after-school club for @whenweplay.org, we only got through Sushi Go and Love Letter (both were enjoyed by at least half the club). This week's agenda: - Flip 7 - Set - Five Crowns (or Skip-bo, though probably five crowns)

🙃 This sucks. I see everyone talking about how AI empowers to build better software. But I haven’t seen better software, I’ve just seen services I use for work start to break more often 😭

"ChatGPT said..." "CoPilot thinks..." STFU. Not they didn't. No they don't. These tools are not people. They don't talk. They don't think. You're interacting with a fancy illusion. Move accordingly.

Not sure I'm emotionally or psychologically set up to deal with the way LLMs are shoehorned into everything in my industry. I lose so much focus every day because I hear about something new that infuriates me. Blockchain annoyed me. LLMs have me so far past that

The obvious lesson here is not to let LLMs anywhere close to anything production (and deeper still... don't use LLMs), BUT that's not what I want to talk about with this article. I'd rather talk about asking an LLM to "explain" itself. A quick 🧵 breakdown www.tomshardware.com/tech-industr...

"Free it up, and watch what happens. Some people will make bad art. Some people will make good art. Some people will make a microtonal double-guitar and paint themselves with polka dots. Most people will just live better, which is also fine." All of that is right, none more than the last sentence

Support your public library. Defend your public library. Slay the enemies of your public library.

Over on Whatever, I use my recent visits to film/TV companies as a framing to note how my not "cognitively offloading" any part of my creative process to "AI" is making my career more secure, not less: whatever.scalzi.com/2026/04/24/a...

Week 2 of running @whenweplay.org's Spring board game program, we're running "Unusual card games" - Love Letter - Sushi Go If we get through those, we'll run - Set - Flip 7

“…all I can tell you is that it is a failure when you ask people to adapt to computers. Computers should adapt to people. Asking people to make themselves more legible to software —to turn themselves into a database — is a doomed idea.”

yesterday meta announced it will lay off 10% of its workforce, the latest tech company to use the "headwinds" of AI adoption to justify mass layoffs. we got into this in the supporters' section of the @404media.co podcast this week:

I did it. I made a song.

“Stop eating out to afford tokens” is a wild thing to say

The soundtrack for the revolution. (and legit has been stuck in my head all morning)

I don't need your "output" to be polished. I need it to be real.

The bill always comes due. When LLMs are used to generate output (any output), the labor is not "done." The labor is just moved to the next person in the chain. Text is either then edited/rewritten (labor for editor) or published as is (labor for the user).

Every 1:1 with my manager

Battle Sheep was the favorite, but one kid loved Kingdomino (and all the kids understood it) and one kid loved Tsuro

Starting a new after-school program for @whenweplay.org today. We'll see what this batch of kids is up for in terms of board games Today, we're teaching Battle Sheep, Tsuro, and Kingdomino. We only taught one of those last semester. Excited to see if the kids enjoy these

GenAI has no place in education. GenAI is, in fact, the antithesis of education. Not to mention, access to these technologies come with potential for addiction, psychological harm, and pure deskilling of the next generation. The only reason to put them in schools is to make money for AI companies.

Easily in my top 3 focaccia loaves I've ever made. Light dusting of extra sharp cheddar on top; mix of a heritage flour with King Arthur flour; mix of commercial yeast and a bit of my sourdough starter. Really really good

OpenAI: We’re burning money like the Joker. A miracle needs to happen for us to turn a profit Microsoft: Please please use our AI systems, we’re teetering on the edge here Anthropic: I wonder what’ll kill us first: lawsuits, regulations or model collapse Media and universities: AI is here to stay